PhD Candidate @ Max Planck Institute for Informatics | Computer Graphics, Computer Vision and Machine Learning
Saarbrücken, Germany
Joined October 2011
- Tweets36
- Following233
- Followers198
- Likes116
Navami Kairanda retweeted
🚀 Excited to kick off the Graphics × Science Workshop at #SIGGRAPH2026!
Computer graphics is becoming a foundational tool for scientific discovery. From computational imaging and molecular modeling to physical simulation, robotics, manufacturing, and AI, graphics is helping us model, understand, and design the physical world.
Looking forward to an exciting program featuring 3 keynote speakers and 58 highlighted papers.
🔗 graphics4science.github.io/2…
#AI4Science #ComputerGraphics #ScientificComputing #Simulation #NVIDIA #NVIDIAOmniverse #NVIDIAAI
Presenting my latest work at #ICLR2026! Stop by the poster #210, Pavillion 3, Rio centro at ⏰ 3:15–5:45 PM to learn more!
🔗 Project page: 4dqv.mpi-inf.mpg.de/DInf-Gri…
🇧🇷 ∂∞-Grid by @NavamiK et al., #ICLR2026 🇧🇷
• Multi-resolution grids + RBF interpolation
• Infinitely differentiable
• Trained using differential equations (DEs) as losses
• Accurate modelling of physical fields
🔗4dqv.mpi-inf.mpg.de/DInf-Gri…
📝arxiv.org/pdf/2601.10715
[1/2]👇
Navami Kairanda retweeted
🎉 LoST: Level of Semantics Tokenization (CVPR 2026)
What if 3D generation didn’t start from coarse geometry, but from semantics?
We introduce LoST, a semantic-first 3D tokenizer that orders tokens by semantic salience, so even very short prefixes can already decode into complete, plausible, recognizable 3D shapes 🧩
Early tokens capture the principal semantics.
Later tokens refine the details.
Result: Generate 3D shapes starting from as few as 1 token!
The goal is to have high fidelity shapes even with very few tokens, with more tokens semantic alignment increases.
Project: lost3d.github.io/
Paper: arxiv.org/abs/2603.17995
@CVPR @AdobeResearch @Adobe
Navami Kairanda retweeted
What can make monocular 4D reconstruction even more ill-posed? Replace RGB inputs with a monocular event stream🥷 Here's the first solution to this new problem at #BMVC2025: arxiv.org/pdf/2510.11717.
Led by Takuya Nakabayashi (Keio🗼), in a team with @NavamiK and @WeLoveHideo.
🎬 It’s #CVPR2025 season!
Meet Thin-Shell-SfT — we enable fine-grained, monocular non-rigid 3D surface tracking with:
✅ Continuous and adaptive deformation fields
✅ Continuous thin-shell physics prior
✅ Surface-induced 3D Gaussian splatting
🧵(1/2)
📍 Catch us at #CVPR2025:
🖼 Poster #68, Exhibit Hall D
🗓 Sat, June 14 | ⏰ 8:30–10:30 AM
Also at:
🧠 1st Workshop on Vision Meets Physics
🗓 Thu, June 12 | ⏰ ~10:50–11:00 AM
🌐 visionmeetphysics.github.io/
🔗 Project page: 4dqv.mpi-inf.mpg.de/ThinShel…
(2/2)
Navami Kairanda retweeted
Can we generate meaningful speech-driven motion by extracting semantics from data?
✅Yes, we can! Check out our latest #CVPR2025 work on generating semantically rich gestures using Diffusion-based RAG 🚀
🎥: vcai.mpi-inf.mpg.de/projects…
@VcaiMpi
#AI #MotionGeneration #EmbodiedAI
Navami Kairanda retweeted
🚀 Check out our #CVPR2025 paper BimArt! 👐🔧
We generate 3D bimanual interactions with articulated objects by:
✅ Predicting contact maps from object trajectories
✅ Using an articulation-aware feature representation
Project Webpage: vcai.mpi-inf.mpg.de/projects…
@VcaiMpi
Join me for the poster session tomorrow (11am -2pm PST, East Exhibit Hall A-C, #3911) to know more about neural implicit cloth simulation!
Let me know if you are at #NeurIPS2024/Vancouver and would like to meet up!
Interested in neural cloth simulation? Don't miss our recent work NeuralClothSim at #Neurips 2024.
You can find @NavamiK at East Exhibit Hall A-C #3911 on Friday 13 Dec 11 a.m. PST — 2 p.m. PST.
Project page: 4dqv.mpi-inf.mpg.de/NeuralCl…
Navami Kairanda retweeted
Interested in neural cloth simulation? Don't miss our recent work NeuralClothSim at #Neurips 2024.
You can find @NavamiK at East Exhibit Hall A-C #3911 on Friday 13 Dec 11 a.m. PST — 2 p.m. PST.
Project page: 4dqv.mpi-inf.mpg.de/NeuralCl…
Navami Kairanda retweeted
NeuralClothSim (continued):
*uses a non-linear anisotropic material model
*allows material interpolation and simulation editing
NDFs:
*allocate the capacity to the deformation details
*allow surface state queries at arbitrary spatial resolutions without retraining
Navami Kairanda retweeted
Existing solvers cannot be easily integrated into neural networks. We rethink physically accurate quasistatic cloth simulation. NeuralClothSim:
*encodes deformations as a neural deformation field (NDF)
*supervises NDF equilibria relying on non-linear Kirchhoff–Love theory
👇
Navami Kairanda retweeted
Can cloth simulation benefit from a neural implicit spatiotemporal deformation field supervised by the Kirchhoff–Love thin plate theory?
Answers: @NavamiK et al. NeuralClothSim. @NeurIPS2024.
Paper: arxiv.org/pdf/2308.12970v3
Project page & Source code: 4dqv.mpi-inf.mpg.de/NeuralCl…
👇
Navami Kairanda retweeted
Check out our #ECCV Relightable Neural Actor! It provides pose and light control, self shadows, surface wrinkles, and material editing at inference time!
Project: people.mpi-inf.mpg.de/~dluvi…
Team: @DiogoLuvizon, @VGolyanik, @AdamKortylewski, @marc_habermann, C. Theobalt
Navami Kairanda retweeted
How can we meta-learn a prior for human radiance and SDF fields to tackle ambiguities in the sparse (few camera) performance capture setting?
Check out our #ECCV 2024 work MetaCap!
Project page: vcai.mpi-inf.mpg.de/projects…
Team: @GuoxingSun, @rishabh_dabral, @FuaPv, Christian
Navami Kairanda retweeted
#ECCV2024 Presenting ReMoS, our latest work on Reactive Motion Synthesis for modeling 2-person interactions.
Project page: vcai.mpi-inf.mpg.de/projects…
Team: @ghosh_ani12, @rishabh_dabral, @VGolyanik, C. Theobalt, P. Slusallek
Navami Kairanda retweeted
Check out our TPAMI 2023 survey paper on visual AIGC and multimodal generation, delving into GANs, diffusion models, autoregressive models, and NeRFs.
Project page: fnzhan.com/Generative-AI/ GitHub: github.com/fnzhan/Generative… Paper: arxiv.org/abs/2112.13592
Navami Kairanda retweeted
𝐃𝐞𝐜𝐚𝐟, please!☕️💆♂️
A work with S. Shimada, P. Pérez and C. Theobalt at @SA2023Sydney 🦘: We are excited to introduce a new approach for non-rigid 3D reconstruction of hands interacting with faces from monocular videos.
vcai.mpi-inf.mpg.de/projects…
#SIGGRAPHAsia2023 #HCI #AR #VR
Navami Kairanda retweeted
Submit your latest works at the intersection of quantum and visual computing to the special issue of IEEE Computer Graphics and Applications!
Submissions due: 22 February 2024
Publication: October 2024
computer.org/digital-library…
please rt
#QuantumComputing
#量子コンピューティング
Navami Kairanda retweeted
Check out @Wanyue__Z's new work titled 𝐑𝐎𝐀𝐌:
• Object-aware 3D motion generation for virtual characters
• Generalises to novel objects of learnt categories
• Works with a single reference object per category in the dataset
arxiv.org/abs/2308.12969
vcai.mpi-inf.mpg.de/projects…